Translational engineering for addressing limitations in leukemia drug screening

Abstract

This thesis is aimed at addressing the limitations of leukemia drug screening platforms and improve their predictive power by employing state-of-the-art translational-engineering approaches. Drug screening has provided important information about tumor-growth inhibition and tumor-cell killing and has driven the increase in cancer survival rates over the last decades, especially for acute lymphoblastic leukemia patients. More recently, the genomic profiling of primary leukemia cells has helped to understand the genetic lesions that occur in leukemia. Despite such progress, the treatment of resistant and relapsed leukemia remains challenging. Recently developed drug-screening platforms have been coupled with molecular disease profiling to provide functional information to select treatment options for high-risk leukemia patients. The clinical benefits of such screening platforms have already been reported, indicating their translational power. However, existing platforms cannot be used to predict the responses to prodrugs, a class of therapeutics that often require metabolic activation to become effective. In addition, existing platforms mostly rely on endpoint measures or few time points to assess drug efficacy and cannot probe real-time drug effects, which would provide access to information on pharmacodynamics of drugs. Microfluidics and advanced 3D cultures offer a powerful toolbox to enable the testing of prodrugs or the real-time screening of drug effects. 3D cell cultures replicate tissue-specific cell-cell or cell-scaffold interactions, e.g., via cell signaling, and promote tissue-specific functions. Microphysiological systems rely on microfluidics to combine different 3D tissue models and to recapitulate organism-level tissue-tissue interactions. In addition, integrated microsensors enable label-free and real-time assessment of the status of the cell cultures. To address the limitations of functional drug screening platforms, three dedicated platforms have been conceived: i. A microphysiological drug screening platform was designed to co-culture primary patient-derived leukemia cells, mesenchymal stromal cells, and human liver microtissues within the same microfluidic platform. The developed platform recapitulated the metabolic activation of ifosfamide in vitro. Sample-specific sensitivities to lethal and short-lived ifosfamide metabolites were identified in primary leukemia, which could not be assessed in existing drug screening assays. ii. Feeder-free cultures of primary patient-derived leukemia cells were realized by supplementing the culture medium with selected cytokines. A high-throughput screening methodology was used to optimize feeder-free media for primary cells derived from three leukemia patients. Observed viabilities in the feeder-free cultures after 3 days were similar to those in state-of-the-art, stroma-based co-cultures. iii. A platform with integrated sensors for impedance cytometry enabled the label-free detection of drug efficacy on suspension cultures of leukemia cells. Cell-line-based experiments demonstrated a basis for a scalable in vitro assay to screen time-dependent drug effects on feeder-free cultures of primary cells. The presented platforms have the potential to be utilized for functional screening, with the goal of further increasing the predictive power of existing drug-screening assays to improve the clinical outcome of patients.